🚀 go-pugleaf

In article <2913@enea.se> sommar@enea.UUCP(Erland Sommarskog) writes:
>Stephen Crawley (scc@cl.cam.ac.uk) writes:
>>2)  Declare the loop body procedure in the same (non-global) scope as
>>    the local variables for the loop.  The loop locals are then available
>>    to the loop body procedure as non-locals.  This assumes that your
>>    language allows this ... C, BCPL, Pascal, Modula-*, Ada etc don't, 
>>    but Algol-68 and Mesa do, as do any languages which support proper
>>    procedure closures.
>> <Example deleted for brevity>
>
>I'm not entirely sure what you mean, so my apologies if my objection
>isn't relevant. But looking at your example, it seems something that
>is straghtforward to do in Ada. 

Arland's objection is both relevant and correct.

Since Ada does not allow you to treat procedures as values, the mechanism
I proposed (passing procedure as parameters to other procedures) does not
work.

However, Ada does allow you to use procedures as parameters to a generic
package.  As Erland pointed out, this allows you to construct iterators.
[It is sad that Ada forces you to go to the trouble of constructing a
 generic package for a data structure you want to iterate over.]

Perhaps I shouldn't make statements about languages I've never used ...

>(There are probably some people objecting strongly against the use
>of the exception here. And I may well agree it's brute force. 
>I included it just to show that there is a way out, ever the author 
>of the iterator forgot to put in a stop condition.)  

I'll excuse you! :-)  End conditions on pseudo-iterators are frequently
messy.  Yet another reason for wanting REAL iterators in a language.

[Sorry about the crap on the end of my last article.  People around
 here have made it unnecessarily hard to post news ... and that was 
 the result.]

-- Steve